30.04.2024

A new Stable Diffusion neural network has been released. It generates pictures according to the description

You can create images directly in the browser. The more complex the query, the better the result. opisaniju-885be24.jpg” alt=”A new Stable Diffusion neural network has been released. It generates pictures according to the description” />

Daria Gromova

Author of Lifehacker

A new one is out Stable Diffusion neural network. It generates images according to the description

The developers of the Stable Diffusion neural network, previously available only in beta version, have posted the project in the public domain. Now anyone can try the service for converting text queries into images.

The easiest way to test Stable Diffusion is on Hugging Face, but there are long queues on the site: one request is processed in about 5 minutes. Through the DreamStudio platform, everything works much faster, but you need an account (you can log in through your Google account or Discord).

In addition, in DreamStudio, each user is given a limited number of credits that can be spent on generation. The more pictures and the higher the resolution, the more expensive each iteration is. To restore previously generated pictures, you also need to spend credits, so you won’t be able to have fun for a long time.

A new one is out Stable Diffusion neural network. It generates images according to the description

The project was also posted on GitHub. For now, it requires a computer with an NVIDIA graphics card and 6.9 GB VRAM, in the future it is planned to add support for AMD video chips, including the Apple M1/M2.

How well does everything work? Depends on the user. The more complex and precise the queries, the more chances for cool results. For example, here are the results for “Daenerys Targaryen in Alphonse Mucha style dragon wedding dress”:

A new Stable Diffusion neural network has been released. It generates images according to the description

This is the “Wizard Yoda in cyberpunk city.

A new one is out neural network Stable Diffusion. It generates pictures according to the description

And this is “Harry Potter horror”. It's scary, but not in the way it should be.

A new Stable Diffusion neural network has been released. It generates pictures according to the description

In other words, you cannot force a neural network to think instead of yourself : for a clear result, you need to think about what you want to get as an output.

Stable Diffusion allows you to adjust the accuracy of matching the request, the number of steps and other parameters. The maximum possible resolution is 1024 × 1024 pixels, up to 9 pictures can be displayed simultaneously. The neural network cannot increase the resolution, for this you will have to rely on third-party programs.

If you try the service, share the results in the comments!

Cover: Lifehacker, Stable Diffusion